58
J. Agric. Food Chem. 1989, 37,58-60
Gresham, T. L.; Jansen, J. E.; Shaver, F. W.; Bankert, R. A,; Beears, W. L.; Prendergast, M. G. @-Propiolactone.VI. Reactions with Phenols, Thiophenols and Their Salts. J . Am. Chem. SOC.1949, 71,661-663. Hayashi, M.; Wada, K.; Munakata, K. New Nematicidal Metabolites from a Fungus, Zrpex lacteus. Agric. Biol. Chem. 1981, 45, 1527-1529. Hayashi, M.; Wada, K.; Munakata, K. Synthesis and Nematicidal Activity of Phenoxypropionic Acid Derivatives. Agric. Biol. Chem. 1983,47, 2653-2655.
Koelsch, C. F. The Identification of Phenols. J. Am. Chem. SOC. 1931,53, 304-305. Lindley, M. G.; Rathbone, E. B. U.K. Patent Application GB2157148A, 1985. Van Straten, S., Maarse, H., Eds. Volatile Compounds in Food: Qualitative Data; Institute CIVO, Analysis TNO: Zeist, The Netherlands, 1983. Received for review November 3,1987. Accepted June 27,1988.
Chirality of 2 4 4-Methoxyphenoxy)propanoicAcid in Roasted Coffee Beans: Analysis of the Methyl Esters by Chiral High-Performance Liquid Chromatography Elner B. Rathbone,*J Ronald W. Butters, Derek Cookson, and Jane L. Robinson2 Methyl 2-(4-methoxyphenoxy)propanoate, previously shown to be present in methylated aqueous extracts of roasted Colombian Arabica coffee beans, was purified by reversed-phase high-performance liquid chromatography (HPLC) with a Chromspher CI8 column (250 X 4.6 mm). The purified extract was analyzed by chiral HPLC with a Chiralcel OK column (250 X 4.6 mm), which provided excellent separation of the two enantiomers of the ester. The methyl 2-(4-methoxyphenoxy)propanoatein the methylated coffee extract was shown to be composed of predominantly (80%) the (S)-(-)enantiomer.
A new flavoring, 2-(4-methoxyphenoxy)propanoicacid, which is the subject of Tate & Lyle PLC’s British Patent Application No. 2157148A and corresponding patent applications worldwide (Lindley and Rathbone, 1985), has recently been identified in aqueous extracts of roasted Colombian Arabica coffee beans (Rathbone et al., 1989). This flavoring (1) has an asymmetric carbon atom and can therefore exist in two enantiomeric forms: (S)-(-) and (R)-(+)(3 and 5, respectively). Sensory evaluation of the individual enantiomers (3 and 5 ) has shown that the important flavoring functionalities (Rathbone et al., 1989) reside solely in the (S)-(-) form, the (E)-(+)enantiomer being functionally inert. To establish which enantiomer of 2-(4-methoxyphenoxy)propanoicacid occurs in roasted coffee beans, the extracts generated by the previous investigation (Rathbone et al., 1989) were methylated, purified by HPLC, and analyzed by chiral HPLC for the presence of the individual enantiomers of methyl 2-(4methoxyphenoxy)propanoate. OCH I
H3C
COzR
1, R = H 2, R = C H 3
ROPC
H
CH3
3, R = H 4, R =CH3
H3C H
COZR
5, R = H 6, R = C H 3
Tate & Lyle Research & Technology, P.O. Box 68, Reading, Berkshire RG6 2BX, U.K. ‘Present address: Sigma Chemical Co. Ltd., Fancy Rd., Poole, Dorset BH17 7NH, U.K. Present address: Pharmaceutical Department, IC1 Pharmaceuticals, Hurdsfield Industrial Estate, Macclesfield, Cheshire SKlO 2NA, U.K.
MATERIALS AND METHODS Materials. ( S )- (-) - 2- (4-Methoxyphenoxy)propanoic acid (3) was prepared by resolution of the racemic acid (1) via its brucine salt (Fredga and Avalaht, 1965): mp 65-66 ‘C; [aI2OD-42.7O (c 0.95, absolute EtOH); lit. (Fredga and Avalaht, 1965) mp 65-66 ‘C; lit. (Fredga and Avalaht, 1965) [aI2OD-43.1’ (c 2.17, absolute EtOH). Methyl ( S )- (-) -2-(4-methoxyphenoxy) propanoate (4) was prepared by treatment of 3 (2.0 g) with methanol-HC1 (1%,6 mL) at 60 ‘C for 0.5 h. CH2Clz(50 mL) was added to the cooled solution and the mixture extracted with saturated aqueous NaHCO, solution (2 X 20 mL). The CH2C12layer was washed with water (2 X 20 mL), dried (Na2S04),and concentrated in vacuo to afford the product (4) as a pale yellow liquid: 2.0 g; -71.4’, [ ( Y ] ~ ~ ~ ~ ~ -282.8’ (c 2.3, acetone); E1 MS, m / z (relative intensity) 210 (M’, 63), 151 (39), 124 (53), 123 (NO),109 (34), 95 (16), 77 (15); ‘H NMR 6 6.82 (4 H, H-2’, 3’, 5’, 6’, unresolved m), 4.67 (1 H, q, H-2, J 2 , 3 = 6.8 Hz), 3.752, 3.747, (2 X 3 H, 2 S, C02CH3, 4’-OCH3), 1.59 (3 H, d, H-3, J3,2 = 6.8 Hz); 13C NMR 6 172.7 (C-1), 154.4 (C-1’), 151.5 (C-49, 116.3 (C-2’, 69, 114.5 (C-3’, 59, 73.4 (ester CH,), 55.4 (4’-OCH3), 52.1 (C-2), 18.5 (C-3). Extracts A-C and (R,S)-(f)-2-(4-methoxyphenoxy)propanoic acid (1) and its methyl ester (2) were obtained as previously described (Rathbone et al., 1989). HPLCgrade solvents were used throughout. Solvent-resistant filters were purchased from Schleicher and Schuell (West Germany). HPLC Analysis. An LKB HPLC system was used, fitted with two 2150 pumps, an LC 2152 controller, and a rapid spectral detector (2140), scanning 190-370 nm. LKB wavescan (2140-250 version 1.08) and Nelson chromatography (2145-200) software was used to process the data on a Wang PC with IBM PC emulation option (version 3.2). Injection was via Rheodyne injector loops (20 or 250 FL). Either of the following columns was used at room temperature. Column A: Chrompak Chromspher C18HPLC column (250 X 4.6 mm). Elution program: (a)
0021-8561/89/ 1437-0058$01.50/0 0 1989 American Chemical Society
J. Agric. Food Chem., Vol. 37, No. 1, 1989 59
(t4ethoxyphenoxy)propanoic Acid in Coffee Beans
initial composition of chloride buffer-methanol (7030, v/v) changing linearly over 25 min to 4555 (v/v), (b) buffer to methanol ratio (4555) maintained for 10 min, (c) composition changed to 20:80 buffer/methanol (v/v) over 1min, and (d) the latter composition maintained for 9 min to the end of the run (run time 45 min). The gradient was changed to achieve the initial composition (7030, v/v) over 1min, which was maintained for a further 9 min prior to the next injection (total cycle time 55 min). Chloride buffer composition: NaCl (5.74 g) dissolved in 0.002 M HC1 and diluted with methanol (10% v/v in the buffer solution). Flow rate: 1mL/min maintained throughout. Column B: Chiralcel OK column (Daicel Chemical Industries Ltd) (250 X 4.6 mm). Eluent: n-hexane-2propanol (90:10, v/v), isocratic. Flow rate: 1 mL/min. The run time was 45 min. GC-MS. GC-MS analysis was performed on a Kratos MS80RF/DS55 mass spectrometer and data system, directly coupled to a nonpolar bonded phase (BP5, 5% methyl phenylsiloxane) fused silica column (50 m X 0.33 mm (i-d.))supplied by Scientific Glass Engineering Ltd. The GC was temperature-programmed 1min a t 100 "C, 16 "C/min to 250 "C, 10 rnin a t 250 "C; helium a t 0.36 kg/cm2 was used as carrier gas. Injection volumes were 1 pL, with splitless injection a t 200 "C. MS scan rate was 1 s/decade, interscan time 1 s, range mlz 30-800. Direct-insertion mass spectra were recorded on the same instrument, operating at 70 eV with a source temperature of 200 "C and using a heated direct-insertion probe. lH and 13C NMR spectra were obtained on a Bruker WM 250 spectrometer, with CDC1, as solvent. Chemical shifts are referenced to the solvent a t 77.0 and 7.24 ppm, respectively, for 13C and 'H, and are reported (ppm) relative to tetramethylsilane. Melting points were measured on a Kofler hot-stage apparatus and are uncorrected. Optical rotations were measured on a Perkin-Elmer 141 polarimeter in a 1-dm cell. Preparation of Extracts. The preparation of extracts A-C from roasted Colombian Arabica coffee beans was described in a previous paper (Rathbone et al., 1989). The remainder of extract A (ca. 700 pL) was evaporated to dryness, methylated (500 pL of 1% methanol-HCl), and combined with extracts B and C. The combined, methylated extract (extract D) was filtered (0.45 pm) and purified by reversed-phase HPLC prior to analysis by chiral HPLC. HPLC Purification. Extract D (total volume 1.0 mL) was chromatographed on column A in five injections (5 X ~ 1 8 pL), 5 and samples of the eluate were collected at 1-min intervals, spanning the elution time of methyl 2(4-methoxyphenoxy)propanoate (2) (ca. 21.5 rnin). The appropriate fractions from the five runs were combined (total volume 5 mL) and extracted with n-hexane-diethyl ether (2:l) (4 X 0.5 mL). The combined organic extract was concentrated in a vacuum desiccator to a volume of 0.5 mL (extract E). Preparation and Purification of Extracts. The extraction of roasted Colombian Arabica coffee beans with hot water and subsequent purification of the extract by solvent extraction and preparative-layer chromatography are described in a previous paper (Rathbone et al., 1989). This work yielded extracts A (an acetone solution containing the free acid 1) and extracts B and C (methylated aliquots of extract A). The use of chiral HPLC to determine the chirality of the natural 2-(4-methoxyphenoxy)propanoic acid was investigated. None of the chiral HPLC columns evaluated gave adequate separation of the enan-
&
. I 1 1 0 10 30 4 I
1
1
1
Time (min)
Figure 1. Chiral HPLC traces (column B): (a) 250 r L of a 2 ppm solution of methyl (R,S)-(f)-2-(4-methoxyphenoxy)propanoate (2) standard, peak at Y identified as the (E+(-) enantiomer; (b) 250 pL of extract E, showing two peaks at X and Y corresponding to those of compound 2. In both figures, the lower traces (i) were monitored at 225 nm whereas the upper traces (ii) are the background-subtractedchromatograms (see text).
tiomers of the acid; however, Chiralcel OK was found to give excellent resolution of the enantiomers of the methyl ester (2). Chiral HPLC analysis was therefore performed on methylated samples of coffee extracts. Combination of extracts B and C with methylated extract A yielded extract D which was purified by reversed-phase HPLC (column A). The combined eluate was extracted with n-hexane-diethyl ether to separate the components of interest from the buffer. The concentrated extract (extract E) was analyzed by GC-MS (to confirm the presence of methyl 2-(4-methoxyphenoxy)propanoate in the purified extract) and by chiral HPLC. RESULTS AND DISCUSSION GC-MS Analysis. The GC trace of extract E, monitored a t m/z 210,123, and the total ion current, showed a peak at the same retention time as compound 2 (10.3 min). The fragmentation pattern of the mass spectrum corresponding to this peak was essentially the same as that of methyl 2-(4-methoxyphenoxy)propanoate,showing characteristic ions a t m / z 210, 151, 123, and 109 and confirming the presence of ester 2 in extract E. Chiral HPLC Analysis. The Chiralcel OK column produced excellent separation of the enantiomeric methyl esters 4 and 6; the retention times were 30.3 and 21.3 min, respectively, which correspond to an enantiomeric separation factor (a)(Bishop et al., 1986) of 1.48 (Figure la). The peaks were identified by injection of a solution of the (S)-(-) enantiomer (4), which produced a single peak at the retention time of the slower of the two enantiomers. This observation also demonstrates that acid 3 does not racemize under the methylation conditions employed. Extract E (250 pL) was analyzed by chiral HPLC. .Background absorption and noise variation were minimized by use of the LKB wavescan arithmetic subtraction routine [ F ( t ) = A(225 nm) - 0.94(245 nm)] in which irrelevant absorptions (245 nm) were subtracted from those
60
J. Agric. Food Chem. 1989,37,60-66
of interest (225 nm). The resulting chromatogram (Figure l b ) showed the presence of both enantiomers 4 and 6 at the same retention times as those in the standard mixture but in unequal amounts, the (S)-(-) enantiomer being present at a higher concentration than the (I?)-(+) ester. The two peaks were integrated with use of Nelson difference and integration programs, which quantified the peak area ratio for the (S)-(-)and (R)-(+)components as ca 4:l. The natural form of 2-(4-methoxyphenoxy)propanoic acid, which occurs in Colombian roasted coffee, analyzed as the methyl ester, is therefore predominantly the (S)-(--) enantiomer. Registry No. 4, 4276-74-8; 5, 117896-90-9; (R)-(+)-2-(4methoxyphenoxy)propanoic acid, 4276-75-9.
LITERATURE CITED Bishop, R.; Hermansson, I.; Jaderlund, B.; Lindgren, G.; Pernow, C. Direct HPLC Resolution of Racemic Drugs. Znt.Lab. 1986, 16,46-53. Lindley, M. G.; Rathbone, E. B. U.K.Patent Application GB 2157148A, 1985. Fredga, A,; Avalaht, I. Studies on Synthetic Growth Substances. XX. Resolution and Absolute Configuration of 4-Methoxyphenoxy-propionic Acid. Ark. Kemi 1965,24, 425-430. Rathbone, E. B.; Patel, G. D.; Butters, R. W.; Cookson, D.; Robinson, J. L. Occurrence of 2-4(Methoxyphenoxy)propanoic Acid in Roasted Coffee Beans: Analysis by Gas-Liquid Chromatography and by High-Performance Liquid Chromatography. J. Agric. Food Chem., 1989, preceding paper in this issue. Received for review November 3,1987. Accepted June 27,1988.
Oat Phenolics: Avenanthramides, Novel Substituted N-Cinnamoylanthranilate Alkaloids from Oat Groats and Hulls’ F. William Collins
Fractionation of methanolic extracts of oat groats and hulls by anion-exchange chromatography revealed the presence of a series of anionic, substituted cinnamic acid conjugates, trivially named avenanthramides. Two-dimensional thin-layer chromatography (TLC) showed groat extracts contain more than 25 distinct avenanthramides, while hull extracts contained about 20. Some 15 were common to both groat and hull preparations. The substances were purified by repeated column chromatography on Sephadex LH-20, using TLC to monitor purity, and crystallized from aqueous acetone. The complete structures of 10 avenanthramides have been elucidated using ‘Hand 13C nuclear magnetic resonance (NMR), mass spectroscopy (MS), ultraviolet absorption spectroscopy (UV), and hydrolytic techniques and confirmed by total synthesis. The physicochemical properties, potential biological activity, and distribution within the oat grain are discussed.
Despite the widespread utilization of oats in the agrifood sector, little work has been done to characterize oat grain phenolic compounds. The structurally dependent involvement of cereal grain phenolics in a wide variety of biochemical and nutritional processes makes a knowledge of the types and amounts of these phenolics extremely valuable. Such information is important in improving the quality of specalized oat-based feed and food ingredients, in developing oat processing technology, in evaluating dietary input data for human and animal nutrition, and as an aid in elucidating mechanisms of disease resistance. An in-depth study of oat grain phenolics is therefore currently being carried out with particular emphasis on the structural elucidation of “bound” forms of the commonly reported hydroxycinnamic acids. The structures and functionality of some of these conjugated phenolics have recently been summarized (Collins, 1986), but a new group of nitrogen-containing constituents, the avenanthramides, have now been detected. The purpose of this Food Research Centre, Agriculture Canada, Central Experimental Farm, Ottawa, Ontario K1A OC6, Canada. Contribution No. 750, Food Research Centre, Agriculture Canada.
paper is to briefly describe their occurrence, isolation, structure, physicochemical properties, and potential biological activity. EXPERIMENTAL SECTION
Melting points were obtained on a Fisher-Johns apparatus and are uncorrected. Chromatographic separations of extracts and synthetic mixtures were carried out on Sephadex LH-20 (Pharmacia, Canada) using volume-calibrated glass columns and gravity-flow isocratic elution. Solvents used for column chromatography were as follows: I, CHC13-cyclohexane-MeOH-acetic acid (50:40:5:5);11, CHC13-cyclohexane-MeOH-acetic acid (50:35:105); 111, CHC1,-cyclohexane-MeOH-acetic acid (50:30:15:5); IV, CHC13-cyclohexane-MeOH-acetic acid (50:25:20:5); V, acetone-H20-acetic acid (30:65:5). Chromatographic mobilities were recorded as relative elution values, K’, equal to the ratio v,/vb, where v, is the observed elution volume and v b is the total volume of the packed gel (i.e., bed volume). When small sample application volumes V, are maintained, relative to v b (i.e., V,/Vb 6 IO%), the K’values were found to be independent of the size of the columns used and the process could be scaled up or down as required. Analytical TLC was performed on silica gel plates (Baker-Flex, 1B2-F, Canadian
0021-8561/89/1437-0060$01.50/0Published 1989 by the American Chemical Society
